Note: Descriptions are shown in the official language in which they were submitted.
CA 03164201 2022-06-08
WO 2021/118461
PCT/SG2020/050725
METHOD AND APPARATUS FOR DETECTING
YAW-TO-WIND ABNORMALITY, AND DEVICE AND
STORAGE MEDIUM THEREOF
TECHNICAL FIELD
[0001] Embodiments of the present disclosure relate to the technical field
of wind power
generation, and in particular, relate to a method and apparatus for detecting
a yaw-to-wind
abnormality, and a device, and a storage medium thereof
BACKGROUND
[0002] Yaw-to-wind misalignment is a common phenomenon in wind turbines.
Problems
.. such as yaw-to-wind misalignment and large yaw errors may not only lead to
a decrease in the
generation capacity of the unit, but also affect stability and safety of units
of the turbines.
[0003] In related arts, yaw deviation is analyzed by a division method.
That is, different
wind angles are divided at certain intervals. Then power curve fitting is
performed in different
angular intervals. The different angular intervals are compared in terms of
performance of the
power curve. An interval where the power curve is optimal is used as a yaw
deviation angle.
[0004] In the above related arts, due to the limitation of the data
volume, fitting the power
curve in each interval may result in no power output in some wind speed
intervals, which may
further lead to a poor fitting effect of the final power curve, thereby making
the final yaw
detection inaccurate.
SUMMARY
[0005] Embodiments of the present disclosure provide a method and
apparatus for
detecting a yaw-to-wind abnormality, and a device and a storage medium thereof
CA 03164201 2022-06-08
WO 2021/118461
PCT/SG2020/050725
[0006] According to a first aspect of the embodiments of the present
disclosure, a method
for detecting a yaw-to-wind abnormality is provided. The method includes:
[0007] acquiring a wind direction deviation angle within a specified time
period, wherein
the wind direction deviation angle is an angle defined between a wind
direction and a central
axis of a nacelle of a wind turbine;
[0008] calculating a power performance index, wherein the power
performance index is a
dimensionless number to characterize power generation performance of the wind
turbine;
[0009] determining an optimal wind direction deviation angle based on the
power
performance index, wherein the optimal wind direction deviation angle is a
wind direction
deviation angle corresponding to optimal power generation performance of the
wind turbine
within the specified time period;
[0010] determining a current wind direction deviation angle according to
probability
distribution of the wind direction deviation angle, wherein the current wind
direction
deviation angle is a wind direction deviation angle with a greatest
probability of occurrence
within the specified time period; and
[0011] determining that the yaw-to-wind abnormality is detected if a
difference between
the optimal wind direction deviation angle and the current wind direction
deviation angle is
greater than a preset threshold.
[0012] According to a second aspect of the embodiments of the present
disclosure, an
apparatus for detecting a yaw-to-wind abnormality is provided. The apparatus
includes:
[0013] a deviation angle acquiring module, configured to acquire a wind
direction
deviation angle within a specified time period, wherein the wind direction
deviation angle is
an angle defined between a wind direction and a central axis of a nacelle of a
wind turbine;
[0014] an index calculating module, configured to calculate a power
performance index,
wherein the power performance index is a dimensionless number to characterize
power
generation performance of the wind turbine;
[0015] a first determining module, configured to determine an optimal
wind direction
deviation angle based on the power performance index, wherein the optimal wind
direction
deviation angle is a wind direction deviation angle corresponding to optimal
power generation
performance of the wind turbine within the specified time period;
2
CA 03164201 2022-06-08
WO 2021/118461
PCT/SG2020/050725
[0016] a second determining module, configured to determine a current
wind direction
deviation angle according to probability distribution of the wind direction
deviation angle,
wherein the current wind direction deviation angle is a wind direction
deviation angle with a
greatest probability of occurrence within the specified time period; and
[0017] a result determining module, configured to determine that a yaw-to-
wind
abnormality is detected if a difference between the optimal wind direction
deviation angle and
the current wind direction deviation angle is greater than a preset threshold.
[0018] According to a third aspect of the embodiments of the present
disclosure, a
computer device is provided. The computer device incudes a processor and a
memory storing
at least one instruction, at least one program, a code set or an instruction
set therein. The at
least one instruction, the at least one program, the code set or the
instruction set, when loaded
and executed by the processor, causes the processor to perform the steps of
the method
described in the first aspect.
[0019] According to a fourth aspect of the embodiments of the present
disclosure, a
non-transitory computer-readable storage medium storing at least one computer
program
therein is provided. The computer program, when loaded and run by a processor,
causes the
processor to perform the steps of the method described in the first aspect.
[0020] The technical solutions according to the embodiments of the
present disclosure
may achieve the following beneficial effects.
[0021] A power performance index is calculated, an optimal wind direction
deviation
angle is determined based on the power performance index, a current wind
direction deviation
angle is determined according to probability distribution of the wind
direction deviation
angle, and then whether a yaw-to-wind abnormality is detected is determined.
Compared with
the related technology in which yaw-to-wind detection is performed through a
power fitting
curve, in the technical solutions according to the embodiments of the present
disclosure, a
power index is directly used, which avoids the problem of poor power curve
fitting effect and
improves the accuracy of yaw-to-wind detection.
[0022] It should be understood that the above general description and the
following
detailed description are only exemplary and explanatory, and cannot limit the
present
disclosure.
3
CA 03164201 2022-06-08
WO 2021/118461
PCT/SG2020/050725
BRIEF DESCRIPTION OF THE DRAWINGS
[0023] The accompanying drawings, which are incorporated in and
constitute a part of the
description, illustrate embodiments of the present invention and, together
with the description
thereof, serve to explain the principles of the present invention.
[0024] FIG. 1 is a schematic diagram of an implementation environment
according to an
exemplary embodiment;
[0025] FIG. 2 is a flowchart of a method for detecting a yaw-to-wind
abnormality
according to an exemplary embodiment;
[0026] FIG. 3 exemplarily shows a schematic diagram of a wind direction
deviation
angle;
[0027] FIG. 4 is a flowchart of another method for detecting a yaw-to-
wind abnormality
according to an exemplary embodiment;
[0028] FIG. 5 exemplarily shows a schematic diagram of a processing
process of raw
data;
[0029] FIG. 6 exemplarily shows a schematic diagram of a training process
of a power
regression model;
[0030] FIG. 7 exemplarily shows a schematic diagram of a the wind
direction deviation
angle being divided into intervals;
[0031] FIG. 8 is a block diagram of an apparatus for detecting a yaw-to-
wind abnormality
according to an exemplary embodiment;
[0032] FIG. 9 is a block diagram of an apparatus for detecting a yaw-to-
wind abnormality
according to another exemplary embodiment;
[0033] FIG. 10 is a structural block diagram of a terminal according to
an embodiment of
the present disclosure;
[0034] FIG. 11 is a schematic structural diagram of a server according to
an embodiment
of the present disclosure.
DETAILED DESCRIPTION
[0035] Exemplary embodiments are described in detail hereinafter, and
examples thereof
4
CA 03164201 2022-06-08
WO 2021/118461
PCT/SG2020/050725
are shown in the drawings. When the following description refers to the
figures, unless
otherwise indicated, the same numbers in different figures represent the same
or similar
elements. The implementations described in the following exemplary embodiments
do not
represent all implementations consistent with the present disclosure. Rather,
they are only
examples of apparatuses and methods consistent with some aspects of the
present disclosure
as detailed in the appended claims.
[0036] FIG. 1 is a schematic diagram of an implementation environment
according to an
exemplary embodiment. The implementation environment may include a wind
turbine 10 and
a computer device 20.
[0037] The wind turbine 10 may include a base, a tower, a nacelle, a hub,
and blades
(wind mills). The base is to provide a foundation of the wind turbine to
maintain the stability
of the wind turbine. The wind center of the wind turbine needs to be installed
at a certain
height, and the tower is a structure that provides this height. The nacelle is
mainly composed
of a glass fiber reinforced plastic nacelle cover, a main frame, a yaw system,
a generator set, a
gear box, a brake, a frequency converter, a weather station, and other
accessories. The hub is
mainly composed of a variable pitch motor, a reducer, a variable pitch control
cabinet, a slip
ring, a lightning protection wire and other components. The blade is mainly
composed of a
blade longitudinal beam, a blade shell made of suitable materials, a blade
root, an embedded
bolt flange, and a blade lightning protection and flashover.
[0038] Operating parameters of the wind turbine 10 may be acquired through
a
supervisory control and data acquisition (SCADA) system. The acquired
operating parameters
may include wind speed, rotor speed, blade pitch angle, yaw angle, active
power, wind turbine
operating status, and the like. The acquisition method may be to dispose a
sensor at a
corresponding position of the wind turbine.
[0039] After the operating parameters of the wind turbine 10 are acquired,
the operating
parameters may be sent to the computer device 20 for yaw-to-wind abnormality
detection.
The computer device 20 refers to an electronic device with data calculation,
processing, and
storage capabilities. The computer device may be, for example, a personal
computer (PC) or a
server. There may be one or more computer devices 20. For example, when there
are two
.. computer devices (such as a first computer device and a second computer
device), the first
5
CA 03164201 2022-06-08
WO 2021/118461
PCT/SG2020/050725
computer device may be a private network server that directly transmits data
with the SCADA
system, and the second computer device may be a public network server. The
first computer
device may further send data to the second computer device.
[0040] Optionally, after the computer device 20 performs yaw-to-wind
abnormality
detection, if it detects a yaw-to-wind abnormality, it may also generate
warning information,
which is used to indicate the yaw-to-wind abnormality, and send the warning
information to a
warning device such that the maintenance personnel can find and correct the
abnormality in
time.
[0041] In related technologies, a yaw-to-wind abnormality being detected
by means of
power curve fitting has the following problems: Due to the limitation of the
data volume,
there will be no power output in some wind speed intervals, resulting in
discontinuity of
power curve fitting, and a complete power curve cannot be obtained; and due to
the
particularity of the work of the wind turbine, the power output is noisy and
the data is dirty,
and it is difficult to perform complete data cleaning and acquire a good power
output point,
resulting in poor effect of power curve fitting. Considering the above
problems, the final yaw
detection is not accurate.
[0042] FIG. 2 is a flowchart of a method for detecting a yaw-to-wind
abnormality
according to an exemplary embodiment. In this embodiment, this method is
mainly applied to
the computer device shown in the embodiment of FIG. 1. The method may include
the
.. following steps (201-205).
[0043] In step 201, a wind direction deviation angle within a specified
time period is
acquired.
[0044] The wind direction deviation angle is an angle defined between a
wind direction
and a central axis of a wind turbine.
[0045] The specified time period may be any period of time, and may be
selected
according to actual conditions, which is not limited in the embodiment of the
present
disclosure.
[0046] In an optional embodiment, all wind direction deviation angles
within the specified
time period may be acquired.
[0047] An energy source of a wind turbine is wind energy captured by a wind
mill of the
6
CA 03164201 2022-06-08
WO 2021/118461
PCT/SG2020/050725
wind turbine. In the case of accurate yawing, a central axis of the wind
turbine is parallel to a
wind direction, an angle therebetween is 0 , the wind energy captured is the
largest, and the
wind turbine has the best power generation performance. However, due to
inevitable errors in
the manufacturing, installation, and debugging processes of the wind vane, the
wind turbine
cannot yaw accurately, resulting in a decrease in the power generation
performance of the
wind turbine. Therefore, the wind turbine needs to yaw based on the wind
direction deviation
angle to improve the power generation performance of the wind turbine.
[0048] FIG. 3 exemplarily shows a schematic diagram of a wind direction
deviation
angle. In the figure, a straight line AOB represents a central axis of a wind
turbine nacelle. A
straight line COD represents a wind mill plane. A straight line EO represents
a wind direction.
A wind direction deviation angle is an angle defined between the central axis
AOB and the
wind direction EO.
[0049] In an optional embodiment, the wind direction deviation angle may
be directly
measured, or may be calculated through a geometric quantitative relationship
between the
wind direction, the central axis value of the nacelle, and the deviation angle
defined between
the wind direction and the nacelle axis. The wind direction deviation angle is
equal to a
difference between the position value of the nacelle and the wind direction.
[0050] In step 202, a power performance index is calculated. The power
performance
index is a dimensionless number to characterize power generation performance
of the wind
turbine.
[0051] The power performance index may be calculated based on operating
parameters of
the wind turbine, such as wind speed, generator speed, and power.
[0052] The power performance index is a dimensionless number to
characterize power
generation performance of the wind turbine. A dimensionless quantity is often
written as a
product or ratio of two dimensional quantities, but the final dimensional
quantities eliminate
each other to obtain a dimensionless quantity. Using dimensionless numbers for
analysis can
effectively simplify the analysis process and reduce complexity.
[0053] In step 203, an optimal wind direction deviation angle is
determined based on the
power performance index.
[0054] After the power performance index is acquired, an optimal wind
direction
7
CA 03164201 2022-06-08
WO 2021/118461
PCT/SG2020/050725
deviation angle may be determined based on the power performance index. The
optimal wind
direction deviation angle is a wind direction deviation angle corresponding to
optimal power
generation performance of the wind turbine within the specified time period.
[0055] In step 204, a current wind direction deviation angle is
determined according to
probability distribution of the wind direction deviation angle.
[0056] The wind direction deviation angle may be statistically collected,
and a current
wind direction deviation angle may be determined according to probability
distribution of the
wind direction deviation angle. The current wind direction deviation angle is
a wind direction
deviation angle with the greatest probability of occurrence within the
specified time period.
[0057] In step 205, if a difference between the optimal wind direction
deviation angle and
the current wind direction deviation angle is greater than a preset threshold,
it is determined
that a yaw-to-wind abnormality is detected.
[0058] Further, a difference between the optimal wind direction deviation
angle and the
current wind direction deviation angle may be acquired. If the difference is
greater than a
preset threshold, it is determined that a yaw-to-wind abnormality is detected.
After that, the
maintenance personnel can perform yaw correction.
[0059] In addition, when the difference between the optimal wind
direction deviation
angle and the current wind direction deviation angle is less than the preset
threshold, it means
that the yaw deviation has no significant influence and no correction is
required.
[0060] The preset threshold may be set by the maintenance personnel
according to actual
conditions, which is not limited in the embodiment of the present disclosure.
[0061] In summary, in the technical solution according to the embodiments
of the present
disclosure, a power performance index is calculated, an optimal wind direction
deviation
angle is determined based on the power performance index, a current wind
direction deviation
angle is determined according to probability distribution of the wind
direction deviation
angle, and then whether a yaw-to-wind abnormality is detected is determined.
Compared with
the related technology in which yaw-to-wind detection is performed through a
power fitting
curve, in the technical solution according to the embodiments of the present
disclosure
directly, a power index is directly used, which avoids the problem of poor
power curve fitting
.. effect and improves the accuracy of yaw-to-wind detection.
8
CA 03164201 2022-06-08
WO 2021/118461
PCT/SG2020/050725
[0062] FIG. 4 is a flowchart of another method for detecting a yaw-to-
wind abnormality
according to an exemplary embodiment. In this embodiment, this method is
mainly applied to
the computer device shown in the embodiment of FIG. 1. The method may include
the
following steps (401-417).
[0063] In step 401, raw data of a wind turbine within a specified time
period is acquired.
[0064] The raw data of the wind turbine includes operating parameters of
the wind
turbine, including a raw output power of the wind turbine. In addition, the
raw data of the
wind turbine also includes environmental data of the wind turbine and
operating data of
adjacent wind turbines.
[0065] The operating data of the wind turbine refers to relevant data
during the operation
of the wind turbine, such as wind vane angle (i.e., the deviation angle
defined between the
wind direction and the nacelle), nacelle position, generator speed, pitch
angle, active power,
gear box oil temperature, and unit operating status. The environmental data of
the wind
turbine refers to characteristic data of an environment where the wind turbine
is installed,
such as turbulence intensity, wind direction, ambient temperature, and wind
speed. The
operating data of adjacent wind turbines refers to operating data of other
wind turbines
adjacent to the wind turbine, such as output power, blade angle, generator
speed, and wind
mill speed of the adjacent wind turbines.
[0066] In step 402, a wind direction deviation angle within the specified
time period is
acquired.
[0067] After the raw data of the wind turbine is acquired, a wind
direction deviation angle
may be further acquired based on the raw data of the wind turbine.
[0068] This step is the same as or similar to step 201 in the embodiment
of FIG. 2, which
is not repeated herein.
[0069] In an optional embodiment, after acquiring the raw data of the wind
turbine, step
403 may be performed.
[0070] In step 403, cleaned data is acquired by cleaning the raw data of
the wind turbine.
[0071] The cleaning is to remove abnormal data in the raw data of the
wind turbine. The
abnormal data may include at least one of: stuck data, overrun data, and data
that does not
conform to physical principles. In addition, the abnormal data may also
include artificially
9
CA 03164201 2022-06-08
WO 2021/118461
PCT/SG2020/050725
interpolated and modified data.
[0072] The stuck data refers to data that has the same value or
fluctuates within a small
range within a period of time. The overrun data refers to data that exceeds a
reasonable range
of generation power of the wind turbine power generation data under certain
wind speed
conditions. The data that does not conform to physical principles may include
icing data. The
icing data refers to data that causes poor wind turbine power generation
performance due to
icing of the wind turbine due to weather. In addition, the data that does not
conform to
physical principles may also include other data, which is not limited in the
embodiment of the
present disclosure.
[0073] FIG. 5 exemplarily shows a schematic diagram of a processing process
of the raw
data. The data cleaning may also refer to the IEC standard. The valid data
criteria formulated
are as follows: (1) the power is greater than 0.1 KW; (2) the pitch angle is
less than 20'; (3)
the generator speed is greater than the cut-in speed; (4) the non-wind speed
is less than 10 m/s
and the pitch angle is greater than 5'; (5) the wind direction angle is less
than 15 and greater
than -15'; (6) non-downtime; (7) non-failure time; and (8) non-wind
curtailment time.
[0074] In an optional embodiment, still referring to FIG. 5, acquiring
the cleaned data by
cleaning the raw data of the wind turbine includes the following operations.
[0075] (1) The cleaned data is acquired by single-point cleaning the raw
data of the wind
turbine.
[0076] In the single-point cleaning, only target data is taken into
consideration in
detecting whether the target data is abnormal. For example, when detecting
whether power
data is abnormal, other related parameters are not taken into consideration,
and only whether
the power data meets the standard is taken into consideration.
[0077] (2) The cleaned data is acquired by multi-point cleaning the raw
data of the wind
turbine to.
[0078] In the multi-point cleaning, both target data and at least one
piece of other data are
taken into consideration in detecting whether the target data is abnormal. For
example, when
detecting whether power data is abnormal, not only whether the power data
meets the
standard is taken into consideration, but also whether the wind speed meets
the standard is
taken into consideration. The power data and the wind speed are combined to
make a
CA 03164201 2022-06-08
WO 2021/118461
PCT/SG2020/050725
comprehensive judgment.
[0079] It should be noted that the cleaning of the raw data of the wind
turbine may be
only the single-point cleaning or only the multi-point cleaning, or both the
single-point
cleaning and the multi-point cleaning, which is not limited in the embodiment
of the present
disclosure.
[0080] In an optional embodiment, still referring to FIG. 5, after
acquiring the cleaned
data by cleaning the raw data of the wind turbine, the method further includes
the following
operations.
[0081] (1) Interpolation is performed on the cleaned data.
[0082] In view that after data cleaning, data incompletion may lead to
inaccurate analysis,
data interpolation may be performed to fill in missing data.
[0083] (2) The cleaned data is corrected.
[0084] In data correction, the wind speed of the wind turbine is
corrected based on the
environmental data of the wind turbine. In view of the influence of
environmental factors,
atmospheric pressure, ambient temperature, air density, and the like may all
cause an impact
on the wind speed, such that data correction may be performed on the cleaned
data to reduce
the impact of environmental factors on the data.
[0085] In an optional embodiment, after acquiring the cleaned data by
cleaning the raw
data of the wind turbine, whether a data volume of the cleaned data is greater
than a preset
data volume may also be determined. If the data volume of the cleaned data is
greater than the
preset data volume, step 404 may be performed. If the data volume of the
cleaned data is less
than the preset data volume, considering that the data volume is too small for
accurate
analysis, no subsequent analysis is performed.
[0086] In step 404, data characteristics of the cleaned data are
extracted.
[0087] After the raw data of the wind turbine is cleaned, data
characteristics of the
cleaned data may also be extracted.
[0088] In step 405, an expected output power of the wind turbine is
acquired based on the
data characteristics of the cleaned data by invoking a power regression model.
[0089] After the data characteristics of the cleaned data are acquired, a
power regression
model may be invoked to calculate an expected output power of the wind
turbine.
11
CA 03164201 2022-06-08
WO 2021/118461
PCT/SG2020/050725
[0090] In an optional embodiment, as shown in FIG. 6, the power
regression model is
trained by the following steps.
[0091] (1) At least one training sample is acquired, which includes the
raw data of the
wind turbine.
[0092] (2) Data characteristics of the training sample are extracted.
[0093] (3) A trained power regression model is acquired by training the
power regression
model with the data characteristics of the training sample, and adjusting the
parameters of the
power regression model.
[0094] After a training sample is acquired, data characteristics of the
training sample may
be extracted, such as characteristics of the operating data of the wind
turbine, characteristics
of the environmental data of the wind turbine, and characteristics of the
operating data of
adjacent wind turbines. After that, characteristic construction is performed
based on the above
data characteristics. The power regression model is trained with the
constructed
characteristics. The parameters of the power regression model are gradually
adjusted and a
trained power regression model is acquired.
[0095] In an optional embodiment, when the power regression model
satisfies a training
stop condition, the training of the model is stopped to obtain a trained power
regression
model. The training stop condition of the power regression model may be
preset, such as a
value of a loss function reaching a preset threshold, the number of training
rounds reaching a
.. preset number of rounds, or a training duration reaching a preset duration,
etc., which is not
limited in the embodiment of the present disclosure.
[0096] In an optional embodiment, after the training is completed, the
trained power
regression model may be tested and verified with a test set. After passing the
verification, the
power regression model is pushed on line such that other devices may invoke it
when needed.
[0097] It should be noted that the power regression model may be a random
forest
regression (RF) model, a gradient boosting decision tree (GBDT) model, or
extreme gradient
boosting (XGBoost) model. In addition, it may also be other models, which is
not limited in
the embodiment of the present disclosure.
[0098] In step 406, a power performance index is acquired based on the
expected output
power of the wind turbine and the raw output power of the wind turbine.
12
CA 03164201 2022-06-08
WO 2021/118461
PCT/SG2020/050725
[0099] The raw data includes the raw output power of the wind turbine.
After the expected
output power of the wind turbine is acquired, a power performance index may be
further
calculated in combination with the raw output power of the wind turbine.
[00100] The power performance index is a dimensionless number to characterize
the power
generation performance of the wind turbine. In the embodiment of the present
disclosure, the
power performance index may be a ratio of the raw output power of the wind
turbine to the
expected output power of the wind turbine. If the power performance index is
greater than 1,
the power generation performance of the wind turbine exceeds the expected
output power of
the wind turbine. If the power performance index is less than 1, the power
generation
performance of the wind turbine does not reach the expected output power of
the wind
turbine.
[00101] In some other embodiments, the power performance index may also be
other
parameters that can express the power performance obtained by mathematical
operations such
as addition, subtraction, multiplication, and division between the raw output
power of the
wind turbine and the expected output power of the wind turbine. The
calculation method may
be freely configured as long as it conforms to physical principles, which is
not limited in the
embodiments of the present disclosure.
[00102] In step 407, the wind direction deviation angle is divided into m
angular intervals,
wherein m is a positive integer greater than 1.
.. [00103] After the wind direction deviation angle is acquired, the angle may
be divided into
intervals. That is, the wind direction deviation angle is divided into m
angular intervals
according to a preset interval, wherein m is a positive integer greater than
1. Each angular
interval includes at least one wind direction deviation angle which
corresponds to the power
performance index one by one.
[00104] FIG. 7 exemplarily shows a schematic diagram of the wind direction
deviation
angle being divided into intervals. Assuming that the wind direction deviation
angle is [-10 ,
1001 and the preset interval is 50, the wind direction deviation angle may be
divided into 4
angular intervals.
[00105] In step 408, an average value of power performance indices of an lth
angular
interval among the m angular intervals is calculated.
13
CA 03164201 2022-06-08
WO 2021/118461
PCT/SG2020/050725
[00106] Since the value of the data and the amount of data in each angular
interval are not
the same, after acquiring multiple power performance indices of each angular
interval, an
average value of the power performance indices of each angular interval may be
calculated
for subsequent analysis.
[00107] In step 409, a confidence interval corresponding to the lth angular
interval is
determined based on the average value and a preset confidence probability.
[00108] After an average value of the power performance of each angular
interval is
acquired, a confidence interval corresponding to each angular interval may be
determined in
combination with a preset confidence probability.
[00109] In an optional embodiment, the confidence interval is determined
through the
student's t distribution. It is assumed that the preset confidence probability
is 95%. When the
average value of the power performance of the lth angular interval is
obtained, a variance of
the power performance of the lth angular interval may also be obtained. In
combination with
the average value and the variance, a degree of freedom table of the t
distribution is queried to
determine a corresponding degree of freedom. An upper boundary value and a
lower
boundary value of a confidence interval with a confidence probability below
95% may be
further calculated, thereby determining a confidence interval corresponding to
the lth angular
interval.
[00110] In step 410, lower boundary values of confidence intervals
corresponding to the m
.. angular intervals are compared.
[00111] By step 409, a confidence interval corresponding to each of the m
angular intervals
may be obtained, and on this basis, a lower boundary value of the confidence
interval
corresponding to each angular interval may be compared.
[00112] In step 411, an angular interval corresponding to a confidence
interval with a
maximum lower boundary value is determined as the optimal angular interval.
[00113] By comparing the lower boundary values of the confidence intervals
corresponding to the m angular intervals, in view of the stability and
reliability of the data, an
angular interval corresponding to a confidence interval with a maximum lower
boundary
value may be determined as the optimal angular interval. The optimal angular
interval is an
.. angular interval which the optimal deviation angle falls within.
14
CA 03164201 2022-06-08
WO 2021/118461
PCT/SG2020/050725
[00114] In step 412, a lower boundary value of the optimal angular interval is
determined
as the optimal deviation angle.
[00115] The optimal angular interval includes an upper boundary value and a
lower
boundary value. After the optimal angular interval is determined, a lower
boundary value of
the optimal angular interval may be determined as the optimal deviation angle.
[00116] In an optional embodiment, assuming that the optimal angular interval
is [0.5 , 1 ),
0.5 may be used as the optimal deviation angle.
[00117] In step 413, a current wind direction deviation angle is determined
according to
probability distribution of the wind direction deviation angle.
[00118] In an optional embodiment, an angular interval with a maximum
probability value
among the m angular intervals may be determined as the current deviation
angle; or, an
angular interval corresponding to a median of the wind direction deviation
angles of the m
angular intervals may be determined as the current deviation angle.
[00119] In an optional embodiment, after determining the current wind
direction deviation
angle according to the probability distribution, step 414 may also be
performed.
[00120] In step 414, a percentage of a data volume of the optimal angular
interval is
calculated.
[00121] After the optimal angular interval is determined, a percentage of a
data volume of
the optimal angular interval may be calculated. The percentage is a ratio of
the data volume of
the optimal angular interval to a data volume of the raw data of the optimal
angular interval.
[00122] If the percentage is less than a preset percentage, the analysis
ends.
[00123] If the percentage is greater than the preset percentage, step 415 may
be performed.
[00124] In step 415, whether a difference between the optimal deviation angle
and the
current deviation angle meets a preset threshold is determined.
[00125] A difference between the optimal wind direction deviation angle and
the current
wind direction deviation angle is acquired. If the difference is greater than
a preset threshold,
it is determined that a yaw-to-wind abnormality is detected. After that, the
maintenance
personnel can perform yaw correction.
[00126] In step 416, warning information is generated.
[00127] After it is determined that a yaw-to-wind abnormality is detected,
warning
CA 03164201 2022-06-08
WO 2021/118461
PCT/SG2020/050725
information may be generated to indicate the yaw-to-wind abnormality.
[00128] In step 417, the warning information is sent to a target device.
[00129] The target device may be a warning device in the implementation
environment as
shown in FIG. 1. The warning device may issue a warning upon receiving the
warning
.. information sent by the computer device, such that the maintenance
personnel can correct the
yaw.
[00130] In summary, in the technical solution according to the embodiments of
the present
disclosure, a power performance index is calculated, an optimal wind direction
deviation
angle is determined based on the power performance index, a current wind
direction deviation
angle is determined according to probability distribution of the wind
direction deviation
angle, and then whether a yaw-to-wind abnormality is detected is determined.
Compared with
the related technology in which yaw-to-wind detection is performed through a
power fitting
curve, in the technical solution according to the embodiments of the present
disclosure, a
power index is directly used, which effectively improves the yaw detection
accuracy of the
wind turbine for timely device maintenance and avoids the continuous drop in
power
generation performance and reduces the loss of power generation. Moreover, the
calculation is
fast, without a lot of data to fit the power curve, and the analysis
efficiency is high.
[00131] In addition, by more in-depth data cleaning, the data cleaning effect
is improved,
and errors caused by poor data cleaning effects may be greatly reduced.
[00132] Hereinafter are apparatus embodiments of the present disclosure for
implementing
the method embodiments of the present disclosure. For details that are not
disclosed in the
apparatus embodiments of the present disclosure, reference may be made to the
method
embodiments of the present disclosure.
[00133] FIG. 8 is a block diagram of an apparatus for detecting a yaw-to-wind
abnormality
according to an exemplary embodiment. The apparatus has a function of
implementing the
method embodiments. The function may be realized by hardware, or by hardware
executing
corresponding software. The apparatus may be a computer device, or may be
disposed on the
computer device. The apparatus 800 may include a deviation angle acquiring
module 810, an
index calculating module 820, a first determining module 830, a second
determining module
840, and a result determining module 850.
16
CA 03164201 2022-06-08
WO 2021/118461
PCT/SG2020/050725
[00134] The deviation angle acquiring module 810 is configured to acquire a
wind
direction deviation angle within a specified time period. The wind direction
deviation angle is
an angle defined between a wind direction and a central axis of a wind
turbine.
[00135] The index calculating module 820 is used to calculate a power
performance index.
The power performance index is a dimensionless number to characterize power
generation
performance of the wind turbine.
[00136] The first determining module 830 is configured to determine an optimal
wind
direction deviation angle based on the power performance index. The optimal
wind direction
deviation angle is a wind direction deviation angle corresponding to optimal
power generation
performance of the wind turbine within the specified time period.
[00137] The second determining module 840 is configured to determine a current
wind
direction deviation angle according to probability distribution of the wind
direction deviation
angle. The current wind direction deviation angle is a wind direction
deviation angle with a
greatest probability of occurrence within the specified time period.
[00138] The result determining module 850 is configured to determine that a
yaw-to-wind
abnormality is detected if a difference between the optimal wind direction
deviation angle and
the current wind direction deviation angle is greater than a preset threshold.
[00139] In summary, in the technical solution according to the embodiments of
the present
disclosure, a power performance index is calculated, an optimal wind direction
deviation
angle is determined based on the power performance index, a current wind
direction deviation
angle is determined according to probability distribution of the wind
direction deviation
angle, and then whether a yaw-to-wind abnormality is detected is determined.
Compared with
the related technology in which yaw-to-wind detection is performed through a
power fitting
curve, the technical solution according to the embodiments of the present
disclosure, a power
index is directly used, which avoids the problem of poor power curve fitting
effect and
improves the accuracy of yaw-to-wind detection.
[00140] In some embodiments, as shown in FIG. 9, the apparatus 800 may further
include a
data acquiring module 860.
[00141] The data acquiring module 860 is configured to acquire raw data of the
wind
turbine. The raw data of the wind turbine includes operating data of the wind
turbine,
17
CA 03164201 2022-06-08
WO 2021/118461
PCT/SG2020/050725
environmental data of the wind turbine, and operating data of adjacent wind
turbines. The
operating data of the wind turbine includes a raw output power of the wind
turbine.
[00142] The index calculating module 820 is configured to extract data
characteristics of
the raw data; acquire an expected output power of the wind turbine based on
the data
characteristics of the raw data of the wind turbine by invoking a power
regression model; and
acquire the power performance index based on the expected output power of the
wind turbine
and the raw output power of the wind turbine.
[00143] In some embodiments, as shown in FIG. 9, the apparatus 800 may further
include a
data cleaning module 870.
[00144] The data cleaning module 870 is configured to acquire cleaned data by
cleaning
the raw data of the wind turbine. The cleaning is to remove abnormal data in
the raw data of
the wind turbine. The abnormal data includes at least one of: stuck data,
overrun data, and
data that does not conform to physical principles.
[00145] The index calculating module 820 is configured to acquire the expected
output
power of the wind turbine based on the data characteristics of the raw data of
the wind turbine
by invoking the power regression model if a data volume of the cleaned data is
greater than a
preset data volume.
[00146] In some embodiments, the data cleaning module 870 is used to acquire
the cleaned
data by single-point cleaning the raw data of the wind turbine, wherein in the
single-point
cleaning, only target data is taken into consideration in detecting whether
the target data is
abnormal; and/or acquire the cleaned data by multi-point cleaning the raw data
of the wind
turbine, wherein in the multi-point cleaning, both target data and at least
one piece of other
data are taken into consideration in detecting whether the target data is
abnormal.
[00147] In some embodiments, as shown in FIG. 9, the apparatus 800 further
includes a
data interpolating module 880 and a data correcting module 890.
[00148] The data interpolating module 880 is configured to fill missing data
by
interpolation on the cleaned data; and/or
[00149] the data correcting module 890 is configured to correct the wind speed
of the wind
turbine based on the environmental parameters of the wind turbine by
correcting the cleaned
data.
18
CA 03164201 2022-06-08
WO 2021/118461
PCT/SG2020/050725
[00150] In some embodiments, the first determining module 830 is configured
to: divide
the wind direction deviation angle into m angular intervals, wherein m is a
positive integer
greater than 1, and each angular interval includes at least one wind direction
deviation angle
one-to-one corresponding to the power performance index; calculate an average
value of
power performance indices of an lth angular interval among the m angular
intervals; determine
a confidence interval corresponding to the lth angular interval based on the
average value and
a preset confidence probability; compare power boundary values of confidence
intervals
corresponding to the m angular intervals; determine an angular interval
corresponding to a
confidence interval with a maximum lower boundary value as an optimal angular
interval,
wherein the optimal angular interval is an angular interval which the optimal
deviation angle
falls within; and determine a lower boundary value of the optimal angular
interval as the
optimal deviation angle.
[00151] In some embodiments, the second determining module 840 is configured
to
determine an angular interval with a maximum probability value among the m
angular
intervals as the current deviation angle; or determine an angular interval
corresponding to a
median of the wind direction deviation angles of the m angular intervals as
the current
deviation angle.
[00152] In some embodiments, as shown in FIG. 9, the apparatus 800 further
includes a
data volume calculating module 900.
[00153] The data volume calculating module 900 is configured to calculate a
percentage of
a data volume of the optimal angular interval.
[00154] The result determining module 850 is configured to determine whether
the
difference between the optimal deviation angle and the current deviation angle
meets the
preset threshold if the percentage is greater than a preset percentage.
[00155] It should be noted that when the apparatus according to the above
embodiment
implements its functions, only the division of the above functional modules is
taken for
illustration. In practice, the above functions may be allocated to different
functional modules
as needed. Namely, the internal structure of the device is divided into
different functional
modules to complete all or part of the functions described above. In addition,
the apparatus
provided in the above embodiments belong to the same concept and as the method
19
CA 03164201 2022-06-08
WO 2021/118461
PCT/SG2020/050725
embodiments, and for the specific implementation process thereof, reference
may be made to
the method embodiments, which is not repeated herein.
[00156] FIG. 10 shows a structural block diagram of a terminal according to an
embodiment of the present disclosure. Generally, the terminal 1000 includes a
processor 1001
and a memory 1002.
[00157] The processor 1001 may include one or more processing cores, such as a
4-core
processor, an 8-core processor, and so on. The processor 1001 may be
implemented in at least
one hardware form of a digital signal processor (DSP), a field programmable
gate array
(FPGA), and a programmable logic array (PLA). The processor 1001 may also
include a main
processor and a coprocessor. The main processor is a processor used to process
data in an
awake state, also called a central processing unit (CPU). The coprocessor is a
low-power
processor used to process data in a standby state. In some embodiments, the
processor 1001
may be integrated with a graphics processing unit (GPU). The GPU is used to
render and
draw content that needs to be displayed on a display screen. In some
embodiments, the
processor 1001 may further include an Artificial Intelligence (Al) processor.
The Al processor
is used to deal with calculation operations related to machine learning.
[00158] The memory 1002 may include one or more computer-readable storage
medium,
which may be non-transitory. The memory 1002 may also include a high-speed
random
access memory and a non-volatile memory, such as one or more magnetic disk
storage
devices and flash memory storage devices. In some embodiments, the non-
transitory
computer-readable storage medium in the memory 1002 is used to store at least
one
instruction. The at least one instruction is used to be executed by the
processor 1001 to
implement the method for detecting a yaw-to-wind abnormality according to the
method
embodiments of the present disclosure.
[00159] In some embodiments, the terminal 1000 may optionally further include
a
peripheral interface 1003 and at least one peripheral. The processor 1001, the
memory 1002,
and the peripheral interface 1003 may be connected by a bus or a signal line.
Each peripheral
may be connected to the peripheral interface 1003 through a bus, a signal
line, or a circuit
board. Specifically, the peripheral may include at least one of a
communication interface
1004, a display screen 1005, an audio circuit 1006, a camera component 1007, a
positioning
CA 03164201 2022-06-08
WO 2021/118461
PCT/SG2020/050725
component 1008, and a power supply 1009.
[00160] Those skilled in the art may understand that the structure shown in
FIG. 10 does
not constitute a limitation on the terminal 1000, and may include more or
fewer components
than shown in the figure, or combine certain components, or adopt different
component
arrangements.
[00161] FIG. 11 shows a schematic structural diagram of a server according to
an
embodiment of the present disclosure. The server is configured to perform the
method for
detecting a yaw-to-wind abnormality provided in the above embodiments.
[00162] Specifically, the server 1100 includes a CPU 1101, a system memory
1104
including a random access memory (RAM) 1102 and a read-only memory (ROM) 1102,
and a
system bus connecting the system memory 1104 and the central processing unit
1101 1105.
The server 1100 also includes a basic input/output (I/O) system 1106 that
helps to transfer
information between various devices in the computer, and a mass storage device
1107 for
storing an operating system 1113, application programs 1114, and other program
modules
1112.
[00163] The basic input/output system 1106 includes a display 1108 for
displaying
information and an input device 1109 such as a mouse and a keyboard for a user
to input
information. The display 1108 and the input device 1109 are both connected to
the central
processing unit 1101 through an input and output controller 1110 connected to
the system bus
1105. The basic input/output system 1106 may further include an input and
output controller
1110 for receiving and processing input from multiple other devices such as a
keyboard, a
mouse, or an electronic stylus. Similarly, the input and output controller
1110 also provides
output to a display screen, a printer, or other types of output devices.
[00164] The mass storage device 1107 is connected to the central processing
unit 1101
through a mass storage controller (not shown) connected to the system bus
1105. The mass
storage device 1107 and its associated computer-readable medium provide non-
volatile
storage for the server 1100. That is, the mass storage device 1107 may include
a
computer-readable medium (not shown) such as a hard disk or a CD-ROM drive.
1001651 Without loss of generality, the computer-readable medium may include
computer
storage medium and communication medium. The computer storage medium includes
volatile
21
CA 03164201 2022-06-08
WO 2021/118461
PCT/SG2020/050725
and nonvolatile, removable and non-removable medium implemented in any method
or
technology for storing information such as computer-readable instructions,
data structures,
program modules or other data. The computer storage medium include a RAM, a
ROM, an
erasable programmable read-only memory (EPROM), an electrically erasable
programmable
.. read-only memory (EEPROM), a flash memory or other solid state storage
technologies, a
CD-ROM, a DVD or other optical storage, a tape cassette, a magnetic tape, a
disk storage or
other magnetic storage devices. Those skilled in the art may know that the
computer storage
medium is not limited to the above. The system memory 1104 and the mass
storage device
1107 may be collectively referred to as memory.
[00166] According to various embodiments of the present disclosure, the server
1100 may
also be connected to a remote computer to run over a network such as the
Internet. That is, the
server 1100 may be connected to a network 1112 by a network interface unit
1111 connected
to the system bus 1105. In other words, the network interface unit 1111 may
also be used to
connect to other types of networks or remote computer systems (not shown).
.. [00167] The memory also includes at least one instruction, at least one
program, a code set,
or an instruction set. The at least one instruction, at least one program,
code set, or instruction
set is stored in the memory and configured to be used by one or more
processors to implement
the method for detecting a yaw-to-wind abnormality.
[00168] In an exemplary embodiment, a computer device is also provided. The
computer
device may be a terminal or a server. The computer device includes a processor
and a
memory. The memory stores at least one instruction, at least one program, a
code set, or an
instruction set. The at least one instruction, the at least one program, the
code set, or the
instruction set, when loaded and executed by the processor, causes the
processor to perform
the method for detecting a yaw-to-wind abnormality.
[00169] In an exemplary embodiment, a computer-readable storage medium is also
provided. The computer-readable storage medium stores at least one computer
program
including at least one instruction, at least one program, a code set, or an
instruction set. The at
least one instruction, the at least one program, the code set, or the
instruction set, when loaded
and executed by a processor, causes the processor to perform the method for
detecting a
yaw-to-wind abnormality.
22
CA 03164201 2022-06-08
WO 2021/118461
PCT/SG2020/050725
[00170] It should be understood that the term "a plurality of' mentioned
herein refers to
two or more, and the term "and/or" describes an association relationship of
associated objects,
indicating that there may be three types of relationships. For example, A
and/or B may
indicate three situations, i.e., A exists alone, A and B exist simultaneously,
and B exists alone.
The character "/" generally indicates that the associated objects are in an
"or" relationship.
[00171] The above are just exemplary embodiments of the present invention,
which will
not limit the present disclosure. Any modifications, equivalent replacements
and
improvements made within the spirits and principles of the present disclosure
shall all fall in
the protection scope of the present disclosure.
23
